From owner-freebsd-multimedia Tue Apr 25 11:49:44 2000 Delivered-To: freebsd-multimedia@freebsd.org Received: from cis.ohio-state.edu (mail.cis.ohio-state.edu [164.107.115.5]) by hub.freebsd.org (Postfix) with ESMTP id EBE5D37B5C3 for ; Tue, 25 Apr 2000 11:49:36 -0700 (PDT) (envelope-from matey@cis.ohio-state.edu) Received: from gamma.cis.ohio-state.edu (matey@gamma.cis.ohio-state.edu [164.107.112.13]) by cis.ohio-state.edu (8.9.1/8.9.1) with ESMTP id OAA08356; Tue, 25 Apr 2000 14:49:35 -0400 (EDT) Received: (from matey@localhost) by gamma.cis.ohio-state.edu (8.9.1/8.9.1) id OAA27702; Tue, 25 Apr 2000 14:49:35 -0400 (EDT) Date: Tue, 25 Apr 2000 14:49:27 -0400 From: Alexander Matey To: Cameron Grant Cc: freebsd-multimedia@FreeBSD.ORG Subject: Re: cvs commit: src/sys/dev/sound/pcm channel.c channel.h dsp.c Message-ID: <20000425144927.A61036@cis.ohio-state.edu> References: <200004231809.LAA27975@freefall.freebsd.org> <39055176.25A5AE68@altavista.net> <20000425113209.A33037@cis.ohio-state.edu> <000701bfaed3$216da6f0$0304020a@NENYA> <20000425130215.A60758@cis.ohio-state.edu> <000901bfaedc$d88d24b0$0304020a@NENYA> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0.1i In-Reply-To: <000901bfaedc$d88d24b0$0304020a@NENYA>; from gandalf@vilnya.demon.co.uk on Tue, Apr 25, 2000 at 06:36:53PM +0100 Sender: owner-freebsd-multimedia@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, Apr 25, 2000 at 06:36:53PM +0100, Cameron Grant wrote: > > i'm not actually opposed to it, i've just not gotten to it yet. i will get > to it eventually - submission of a patch including MAKEDEV would speed that > up. Currently MAKEDEV snd* creates only one channel per each sound device. Unless we won't to change that and create device nodes for several channels, there's nothing in MAKEDEV that has to be changed. I'm resending you in separate email (there's no point to propagate it on the maillist) the patch for sound.c, which I've already sent you earlier. It moves channel number from 2nd byte of device minor to 3rd one. It's trivial. Also, it updates your "PROPOSAL" accordingly. Make sure to review it. > my priorities at the moment are getting the generic layer OSS emulation > right and fixing recording which is just plain broken at the moment. I noticed that chn_clearbuf() clears buffer bottom up which may cause problems if driver's getptr() function returns values that are not DMA_ALIGN_THRESHOLD aligned. This way 1 to 3 bytes can be lost on recording. But I was looking at it a week or so ago and there were commits, so this may not be true now. > > number. Again, this is not going to work since contents of the 2nd byte > are > > going to be overwritten by the kernel deep inside make_dev(). > > the way device numbers are treated by ther kernel has changed since that was > written. Another reason to catch up. -- lx To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-multimedia" in the body of the message